Accurate risk prioritizing is directly related to the effectiveness of risk management. To overcome the shortage of the single numerical evaluation value, aiming at improving the accuracy of risk factors, a new risk priority method was proposed based on geometric characteristics of triangular fuzzy number and Analytic Hierarchy Process (AHP). This method was established on the basis of the fuzzy description of risk factors from experts, after the risk evaluation system was established. Then the fuzzy description of risk was processed with AHP, and fuzzy weights of risk factors were obtained and calculated it by using the geometric characteristics of triangular fuzzy number. Finally, the detailed ranking of risk factors by severity, probability and detection of risk was obtained. The risk priority of forklift system was processed to analyze the feasibility of this method. The results showed that the new method proposed here could fully get the implicit information, such as the weights of experts and the factors. It could improve the rationality of the risk sorting and revise the offset of the result using the traditional method.